Synthesis Analysis of Multi-dimensional Ozone measurements in Coastal Environments Toward Improving Simulations and Advancing Satellite Products
In order to improve the comprehension and forecasting of spatio-temporal variations of surface ozone, it is essential to combine chemical transport model (CTM) simulations with measurement data. The purpose of this thesis is to investigate air quality variability in coastal regions using the cross-validation of CTMs, high quality multi-dimensional lidar data, surface measurements, and satellite retrievals. This study first evaluates diurnal ozone patterns in the Houston-Brazoria- Galveston (HGB) region using a clustering method to better understand meteorological patterns (Bermuda High and Low-level jet) and their interaction with ozone variability. Results demonstrated the ability of the clustering method to group ozone variability more accurately than a simple method tested and showed clear meteorological influence based on the different clusters. Next, we utilize multi-dimensional measurements from ozone lidar in conjunction with both an offline GEOS-Chem CTM simulation and the online GEOS-Chem simulation, GEOS- CF, to investigate spatio-temporal variations of coastal ozone during three air quality campaigns: 2017 Ozone Water-Land Environmental Transition Study (OWLETS)-1, 2018 OWLETS-2, and 2018 Long Island Sound Tropospheric Ozone Study (LISTOS). The cluster results effectively capture and emphasize the diverse temporal and vertical variations observed in multiple cases throughout the campaign periods. Results indicate both models struggle to simulate mid-level ozone (2000 - 4000 m) and a generally weak agreement to the lidar (R = 0.12 and 0.22, respectively). Both models have a good agreement (R ≈ 0.70) in the low-level altitude range (0 - 2000 m). Finally, we investigate regional impacts of anthropogenic emissions on tropospheric ozone during a recent campaign, Tracking Aerosol Convection ExpeRiment – Air Quality (TRACER-AQ) using the Weather Research and Forecasting (WRF) model coupled with Chemistry (WRF-Chem) at a fine resolution (4 km), ozone lidar observations, surface monitoring stations, and Tropospheric Monitoring Instrument (TROPOMI) nitrogen dioxide vertical column densities. Results reveal the model is better able to capture mid-level/free tropospheric ozone and underestimates surface/boundary layer ozone during two episodes. The results also reveal an underestimation of simulated nitrogen dioxide levels in urban areas, as well as over the Galveston Bay/near the Gulf coast, potentially attributable to discrepancies in modeled emission sources.